[3dem] IHRSR++ and SPIDER 21.02

Morgan, David Gene dagmorga at indiana.edu
Sun Mar 6 11:53:20 PST 2016


Hi,


    I couldn't agree with Ed more.  Although selection rules may initially seem to make the difficult concepts behind helical diffraction (and the image processing that makes use of these concepts) somewhat easier to swallow, in the long run, they cause more trouble than they are worth.  I have found the formalism of DeRosier's (n,Z) net and subunit rise and rotation to be much more satisfactory, and often use the example of a helix built using a subunit rotation of e degrees and a subunit rise of pi Angstroms.  Yes, there are integers that would closely approximate this structure, but that description also loses some of it's beauty!


--
            David Gene Morgan
        Electron Microscopy Center
             047D Simon Hall
             IU Bloomington
          812 856 1457 (office)
          812 856 3221 (3200)
      http://iubemcenter.indiana.edu
________________________________
From: 3dem <3dem-bounces at ncmir.ucsd.edu> on behalf of Edward Egelman <egelman at virginia.edu>
Sent: Sunday, March 6, 2016 2:31 PM
To: Hernando J Sosa; Smith, Phillip R.; 3dem at ncmir.ucsd.edu
Subject: Re: [3dem] IHRSR++ and SPIDER 21.02

It depends upon how "classical" you mean! This excellent handout from David DeRosier predates single particle approaches, and from the references must have been written between 1992 and 2000:

http://www.biomachina.org/courses/structures/download/derosier_handout_02.pdf

Since David (with Aaron Klug) was the first to use Fourier-Bessel methods in EM (and publish the first 3D reconstruction, he has some authority in this area. The selection rule is introduced in his handout as a parenthetical alternate description, but the entire Fourier-Bessel formalism presented is in terms of n,Z (where Z is a real number) and not n,l (where l is an integer and arises from an integral selection rule, with u and t being integers). So one does not need (and is hindered by, in my humble opinion) a selection rule to understand and use the Fourier-Bessel approach. Perhaps I was being picky, but u/t does not "select" anything, while the equation l=tn +um selects what Bessel order n can appear on a layer line l. That is why this equation is called the selection rule. Of course, any real number can be represented to good approximation as a ratio of two integers, so if some "classical" program requires integers for u and t, one can always approximate them. I may also suffer from a lack of visual imagination (with regards to Esther's point), as I do not see how saying that a filament twist may range from 13 subunits in 6 turns to 1299 subunits in 600 turns is helpful (a change from 166.15 to 166.28 degrees). I would much rather think in terms of angles. But perhaps that is just my problem.
Regards,
Ed


On 3/6/16 2:07 PM, Hernando J Sosa wrote:

To clear up definitions  in my book describing a helix as  u = X subunits  in  t =  Y turns  (per  axial repeat)  IS the same as providing the selection rule, as these are the only two variables in the selection rule equation  l=tn + um that fully describe the helix.  With these parameters (u & t) the possible Bessel orders for any layer line number are determined according to the selection rule equation.   The parameters u and t can be simply converted to the alternative way to describe a helix with the twist (phi) angle and rise (h) per subunit as:  h=axial_repeat/u,    phi=360*t/u.



As discussed the problem with this selection rule definition is that it arbitrarily assumes an integer number of turns and subunits per axial repeat.  However, in practice good approximations can be found assuming a long enough axial repeat or using a more general selection rule expressed in function of the twist and rise (e.g. Moody MF. 1990).



When trying to determine the symmetry of an unknown helical specimen (or modifications of a known one) I find it useful to go back and forth between these equivalent definitions so I wouldn't consider unreasonable to talk about selection rules.    Knowing the 'classical' selection would also be necessary if for any reason you want to use 'classical' 3D Fourier-Bessel helical 3D reconstruction packages.





Best



Hernando

________________________________
From: 3dem [3dem-bounces at ncmir.ucsd.edu<mailto:3dem-bounces at ncmir.ucsd.edu>] on behalf of Edward Egelman [egelman at virginia.edu<mailto:egelman at virginia.edu>]
Sent: Saturday, March 05, 2016 10:36 AM
To: Smith, Phillip R.; 3dem at ncmir.ucsd.edu<mailto:3dem at ncmir.ucsd.edu>
Subject: Re: [3dem] IHRSR++ and SPIDER 21.02

No reasonable person would use selection rules any more. They were formulated in the 1950s and arise from a crystallographic-type formulation where a helix is described by the ratio of integers (units/turn or u/t). For real helices, the best description is given by two real numbers, a rise (Angstroms) and a rotation (degrees). The description of those tubes (I assume) is given in Parent et al., Physical Biology:

doi:10.1088/1478-3975/7/4/045004

Regards,
Ed

On 3/5/16 9:49 AM, Smith, Phillip R. wrote:

The data and tutorial that you point to is indeed excellent and a nice testbed for software.

But it would be a huge help if someone could provide the selection rule for the F170A tubes in the data provided, p8:

"The values for the symmetry parameters ([Cn], [rise], [deltaphi]) were derived from the diffraction pattern (derivation not shown).”

Hope you can help…

Very best to all!

-Ross Smith-



On Feb 29, 2016, at 4:39 PM, Edward Egelman <egelman at virginia.edu><mailto:egelman at virginia.edu> wrote:

Hi,
  Unfortunately, there are no good tutorials. Also, the more that I learn the more I realize that it is not as simple as I originally assumed. I would suggest reading three papers as a start:

Egelman, E.H. (2010), “Reconstruction of Helical Filaments and Tubes”, Methods in Enzymology 482, 167-183.

Egelman, E.H. (2014). “Ambiguities in helical reconstruction”. eLife 3:e04969 doi:10.7554/eLife.04969.

Egelman, E.H. (2015). “Three-dimensional reconstruction of helical polymers”, Archives of Biochemistry and Biophysics 581, 54-58.

Regards,
Ed


On 2/29/16 2:58 PM, Johannes Haataja wrote:


Dear all,
        thank you for the replies. I now have an older version of spider.

Regarding IHRSR Prof. Egelman - what would the recommended way/tutorial
for learning to use IHRSR?

My best,
 - J.

P.S. I guess ideally one would just read an article about the theory and
unix/linux man-pages of relevant command line tools and then inductively
reason how one must proceed to apply the method to the problem at hand.
Since I lack such a tenacity, I usually look for tutorials in order to
understand how the softwares/black boxes work. Also, I imagine that for
helical reconstruction, like for any inverse problem, there are many
different methods for recovering the quantit(y/ies) of interest and that
people usually are hesitant to openly aside with particular approach may
it be the right one or obviously the wrong one (e.g. Bayesian vs.
Frequentist interpretation of statistics) ;).



ma, 2016-02-29 kello 12:13 -0500, Michael Radermacher kirjoitti:



I would contact the people in Albany and also
discuss with them the problem you are having
with your version.

Michael

On 2/29/2016 11:46 AM, Johannes Haataja wrote:



Hi,
        does anyone know where to obtain old versions of SPIDER, namely v.
21.02? The oldest from download page is 21.11. The reason for asking is
that I need and older SPIDER version to test IHRSR++ v. 1.5 tutorial


http://cryoem.ucsd.edu/wikis/software/start.php?id=ihrsr


to exclude the possibility that the errors I run into (in the final
reconstruction step) have something to do with SPIDER version.

My best,
        - J.


------------------------------------------------------------
This email message, including any attachments, is for the sole use of the intended recipient(s) and may contain information that is proprietary, confidential, and exempt from disclosure under applicable law. Any unauthorized review, use, disclosure, or distribution is prohibited. If you have received this email in error please notify the sender by return email and delete the original message. Please note, the recipient should check this email and any attachments for the presence of viruses. The organization accepts no liability for any damage caused by any virus transmitted by this email.
=================================


--



Edward H. Egelman, Ph.D.
Harrison Distinguished Professor
Dept. of Biochemistry and Molecular Genetics
University of Virginia

phone: 434-924-8210
fax: 434-924-5069
egelman at virginia.edu<mailto:egelman at virginia.edu>
http://www.people.virginia.edu/~ehe2n<http://www.people.virginia.edu/%7Eehe2n>


--



Edward H. Egelman, Ph.D.
Harrison Distinguished Professor
Dept. of Biochemistry and Molecular Genetics
University of Virginia

phone: 434-924-8210
fax: 434-924-5069
egelman at virginia.edu<mailto:egelman at virginia.edu>
<http://www.people.virginia.edu/%7Eehe2n>http://www.people.virginia.edu/~ehe2n

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.ncmir.ucsd.edu/pipermail/3dem/attachments/20160306/1583c71c/attachment-0001.html>


More information about the 3dem mailing list